John Saunders <johns@rd.scitec.com.au>
Endre Balint Nagy <bne@CareNet.hu>
Geoff Keating <Geoff.Keating@anu.edu.au>
+ Arjan de Vet <Arjan.deVet@adv.IAEhv.nl>
Development of this caching software is funded by the National Science
and a single max object size?
- Don't do TTL calculation whn object is fetched. Instead
do refresh time calculation when object is requested.
- Maybe we can lose StoreEntry->expires?
- lose 'max_expire_age' business?
- - Fix up 'cache_mem' and accounted memory. Divide accounted
- memory into different buckets and have separate config value
- for each?
- Write Perl script to convert squid-1.0 store to squid-1.1
structure.
- Fixed lots of function prototypes, etc (Ed Knowles).
- Added multicast patch (Martin Hamilton).
+ - Replaced 'struct hostent' with 'struct ipcache_addrs' in IP cache.
+ - Added ipcacheCycleAddrs() to round-robin IP addresses.
+ - Added ipcacheRemoveBadAddr() to remove addresses from failed connect()'s.
+ - Changed comm_connect() to comm_nbconnect() and removed other nonblocking
+ connection handling code from other modules.
+ - Improved cache memory usage. Now cache_mem specifies the size
+ of the in-memory data pool. Hot objects use whatever space is
+ not used by in-transit objects.
+ - Removed 'max_hot_object_size' config option.
+ - Fixed virtual-host coredump bug (Aaron Hopkins).
Changes to squid-1.1.beta5 (October 7, 1996):